Dv

Qa

b

c

a

ARRAA

KBSPA

1

trfamao

bmwA(M[tbpibl

0d

Electrochimica Acta 55 (2010) 4717–4721

Contents lists available at ScienceDirect

Electrochimica Acta

journa l homepage: www.e lsev ier .com/ locate /e lec tac ta

etermination of trace selenium by differential pulse adsorptive strippingoltammetry at a bismuth film electrode

ing Zhanga, Xiangjun Lib,∗, Hui Shic, Hongzhouc, Zhuobin Yuanb

Chinese Academy of Inspection and Quarantine, Beijing 100123, ChinaCollege of Chemistry and Chemical Engineering, Graduate University, Chinese Academy of Sciences, Beijing 100049, ChinaDepartment of Chemistry, Capital Normal University, Beijing 100048 China

r t i c l e i n f o

rticle history:eceived 5 January 2010eceived in revised form 24 March 2010

a b s t r a c t

The differential pulse adsorptive stripping voltammetric behavior of selenium (IV)–p-aminobenzene sul-fonic acid–cetyltrimethylammonium bromide system at a bismuth-coated glassy carbon electrode (BiFE)has been investigated. A well-defined and sensitive stripping peak of the selenium (IV)–p-aminobenzene

ccepted 24 March 2010vailable online 2 April 2010

eywords:ismuth film electrodeelenium

sulfonic acid complex was observed at −0.76 V (vs. SCE) in a 0.15 mol/L acetate solution (pH 2.9) at adeposition potential of −0.40 V (for 120 s). The linear range was 2–30 �g/L and the detection limit foran accumulation time of 300 s was 0.1 �g/L. This method was applied to determine the trace amount ofselenium in the samples.

© 2010 Elsevier Ltd. All rights reserved.

-aminobenzene sulfonic acid (ABSA)dsorptive stripping voltammetry

. Introduction

Selenium is an essential trace element, and the toxic and essen-ial levels of this element differ only slightly. Selenium has beeneported to have an anti-cancer effect, protecting the human bodyrom free radicals, and the recommended intake is between 50nd 200 �g/day. Therefore, it is contained in nutritional supple-ents in combination with other trace elements and vitamins or

mino acids. However, it is toxic for humans as well as for marinerganisms [1,2] in higher concentrations.

For the analysis of selenium, several analytical approaches haveeen developed; the total concentration of selenium can be deter-ined by instrumental neutron activation analysis (INAA) [3] asell as by hydride generation atomic absorption spectrometry (HG-AS), inductively coupled plasma atomic emission spectrometry

ICP-AES), inductively coupled plasma mass spectrometry (ICP-S), or electrothermal atomic absorption spectrometry (ET-AAS)

4–7]. Moreover, alternative techniques, such as capillary elec-rophoresis (CE) [8] and chromatography coupled with MS, haveeen reported [9,10]. Voltammetric techniques, however, are inex-

ensive, sensitive and selective. In particular, stripping analysis

s the most selective electroanalytical technique because it com-ines low detection limits (DLs), multielement capabilities and

ow cost [11,12]. Also, in some cases, the preconcentration step

∗ Corresponding author.E-mail address: lixiangj@gucas.ac.cn (X. Li).

013-4686/$ – see front matter © 2010 Elsevier Ltd. All rights reserved.oi:10.1016/j.electacta.2010.03.068

increases the sensitivity, so that the DL can decrease to 100 timeslower than that of direct voltammetric methods, such as polarog-raphy [13]. In most stripping analysis methods, a mercury-basedelectrode is used as the working electrode because of its uniqueability to preconcentrate target metals during the accumulationstep. However, despite the excellent performance of mercury elec-trodes, future regulations and occupational health considerationsmay severely restrict (or even ban) the use of mercury as anelectrode material because of its high toxicity [14]. New alterna-tive electrode materials are, therefore, highly desired to develop“environmentally-friendly” stripping sensors suitable for on-sitemonitoring of heavy metals. Numerous solid electrode materi-als, including gold [15], carbon-based materials [16] and iridium[17], have been tested, but their overall performance has neverapproached that of mercury. Recently, Wang et al. [18,19], Huttonet al. [20], Królicka et al. [21] and Economou and co-workers [22]introduced the bismuth film electrode (BiFE), which is preparedby electroplating a thin layer of bismuth onto a glassy carbon orplatinum substrate, for use in stripping analysis of traces of heavymetals. The behavior of the bismuth film electrode was shown tocompare favorably with that of mercury electrodes, with its attrac-tive properties including high sensitivity, well-defined strippingsignals, good resolution of neighboring peaks (e.g., Cd, Pb, and Zn),

large cathodic potential range, and insensitivity to dissolved oxy-gen. In contrast with mercury electrodes, this last characteristic isan essential property for on-site monitoring. In addition, bismuthis a more “environmentally-friendly” material with a low toxicityand is widely used in medicine and cosmetics.

4 mica A

aattaBmttan(aptwdaa

2

2

RMfia3m

mswpHq

2

awAtBttctce5te

sde1we1

range from 0 to 0.08 mol/L, and remained unchanged in the rangefrom 0.08 to 0.2 mol/L, above this range the peak currents started todecrease. Therefore, the concentration 0.15 mol/L was used in theexperiment.

718 Q. Zhang et al. / Electrochi

Trace quantities of selenium can be determined by means ofdsorptive stripping voltammetry method at mercury, platinumnd gold based electrodes [23–26]. These approaches are based onhe formation of a suitable surface-active metal chelate, its adsorp-ive preconcentration onto the surface of the working electrode,nd the voltammetry measurement of the surface-bound species.asically, the adsorptive cathodic stripping voltammetry (ACSV)ethod consists of two steps: preconcentration or deposition of

he analyte at the electrode surface, and stripping the preconcen-rated analyte at the electrode surface by reduction. In this work, annalytical technique was developed for the determination of sele-ium based on differential pulse cathodic stripping voltammetryDPCSV) using a bismuth film electrode as the working electrodend p-aminobenzene sulfonic acid (abbreviated as ABSA in ouraper) as a complexing agent. The optimization of the developedechnique was achieved by testing unconventional parameters asell as varying the standard voltammetric parameters such aseposition time, deposition potential, and electrolyte acidity. Thepplicability of the whole procedure was applied to the real samplenalysis.

. Experimental

.1. Apparatus and reagents

A Par 2273 electrochemistry workstation (Princeton Appliedesearch Company, USA) and a microwave digestion system (CEModel MDS-2000) were used. The working electrode was a bismuth

lm electrode, the reference was a saturated calomel electrodend the auxiliary electrode was platinum. A magnetic stirrer (ca.00 rpm) was used for convection in the solution during the accu-ulation period.All reagents used were of analytical grade unless otherwise

entioned. The selenium standard solutions were made from atock of 1000 �g/mL (China Iron & Steel Research Institute), dilutedith 10% HCl as required. The 100 mg/L bismuth solution was pre-ared by dissolving the proper amount of Bi(NO3)3·5H2O in 1 mol/LAc–NaAc (pH 4.5). Aqueous solutions were prepared using high-uality water (MilliQ).

.2. Procedure

Prior to use, the glass carbon disk electrode was polished with0.3–0.05 �m alumina slurry on a felt pad, then thoroughly rinsedith water, and sonicated in HNO3, ethanol, and distilled water.fter being cleaned, a bismuth film was potentiostatically elec-

rodeposited on a GCE by applying −1.0 V for 5 min in a 100 mg/Li3+ and 1 mol/L HAc–NaAc aqueous solution. Then the three elec-rodes were immersed into 20 mL of 0.15 mol/L HAc–NaAc solutionhat contained aliquots of Se(IV), p-aminobenzene sulfonic acid andetyltrimethylammonium bromide (CETAB). The preconcentrationime was 120 s with stirring and another 10 s without stirring, thenathodic stripping voltammetry (CSV) was performed in the differ-ntial pulse mode with the following parameters: pulse amplitude,0 mV and time pulse, 0.006 s; voltage step, 5 mV and voltage stepime, 0.1 s; sweep rate, 50 mV/s; before each measurement, thelectrode was cleaned at −1.3 V for 10 s.

The real sample analysis procedure was as follows: the 0.5 gample, 5 mL HNO3 and 1 mL H2O2 were placed into a microwaveigestion tank over night and then digested according to a differ-

nt microwave digestion procedure (3 min at 120 ◦C or 30 min at40 ◦C). When the microwave digestion finished, the digestion tankas cooled down naturally. Then, the digestion tank was heated to

mpty it of HNO3. In order to make sure Se(VI) reduced to Se(IV),mL of 6 mol/L HCl was added into the digestion tank, and the

cta 55 (2010) 4717–4721

mixture was heated in a water bath at 100 ◦C for 15 min [1,27].Then the digestion tank was cooled down to room temperature anddiluted to 10 mL in a volumetric flask with double-distilled waterfor analysis by CSV as mentioned above.

3. Results and discussion

3.1. The selective of the optimum experimental conditions

3.1.1. The effect of the supporting electrolyte and pHBefore the determination of the optimum chemical and elec-

trochemical conditions for the analysis of selenium, preliminaryexperiments were conducted to evaluate the suitability of the elec-trode. The stripping behavior of the Se(IV)–ABSA complex at thebismuth film was determined in different supporting electrolytes,including HAc, HAc–NaAc, HNO3, NH3–NH4Cl, Na2CO3–NaHCO3and others. It was apparent that the HAc and HAc–NaAc solution aremore suitable for the stripping of the Se(IV)–ABSA complex, due tothe favorable stripping peak current, low background current andhigh sensitivity. At a potential of −0.76 V, a sensitive stripping peakof the Se(IV)–ABSA complex was observed (as shown in Fig. 1, curvec).

In our experiment, the effect of pH on the stripping currentof Se(IV)–ABSA complex was determined in 0.15 mol/L HAc andHAc–NaAc supporting electrolytes at different pH values, whereeach solution contained 10 �g/L Se(IV). As is shown in Fig. 2, thepeak currents decrease with increasing pH from 2.9 to 4.7 and thestripping peak current of Se(IV)–ABSA complex in the HAc–NaAcsolution reached the maximum at pH 2.9. When pH > 5, the strip-ping peak is inconspicuous. Therefore, the optimum condition ofthe supporting electrolyte is a pH 2.9 HAc–NaAc solution. The peakpotential shifted negatively with increased pH and the value of�E/�pH is −61 mV, indicating that the reaction at the electrodewas performed in the presence of hydrogen ions.

3.1.2. The effect of HAc–NaAc concentrationThe concentration of HAc–NaAc was varied from 0 to 0.5 mol/L,

and the concentration of Se(IV) was kept at 10 �g/L Se(IV). Thepeak currents increased with the concentration of HAc–NaAc in the

Fig. 1. The effect of ABSA on the stripping current of Se(IV). (a) HAc–NaAc + ABSA(1.0 × 10−4 mol/L), (b) HAc–NaAc + Se(IV) (10 �g/L), and (c) b + ABSA.

Q. Zhang et al. / Electrochimica Acta 55 (2010) 4717–4721 4719

FrH

3

Stwsifa2pbur

2

S

S

3

sda

ig. 2. The effect of pH of the supporting electrolyte on the stripping cur-ent (�) and the peak potential (�) of Se(IV)–ABSA complex in 0.15 mol/LAc–NaAc + 1.37 × 10−5 mol/L CETAB. Se concentration: 10 �g/L

.1.3. The effect of ABSA concentrationIt is reported that Se(IV) could react with ABSA to form a 1:2

e(IV)–ABSA complex [28]. This complex has a stronger adsorp-ive property than Se(IV), so it results in a higher stripping peak,hich can be seen in Fig. 1. When no ABSA is added, there is only a

mall reductive peak from the reduction of Se(IV), but when ABSAs added, the peak shifts positively and the peak current increasesrom the formation of the Se(IV)–ABSA complex and its strongerdsorption. When the concentration of ABSA is varied from 0 to× 10−4 mol/L at 10 �g/L Se(IV) (as shown in the inlet in Fig. 2), theeak currents increase with ABSA until 8 × 10−5 mol/L and thenecome stable. Therefore, the concentration of 8 × 10−5 mol/L wassed in the experiment. According to Sun et al. [29], the electrodeeactions should be as follows:

ABSA + H3SeO3+2H+ → ABSA–Se–ABSA + 3H2O

e(IV)(ABSA)2+Bi → Se(IV)(ABSA)2(ads)(Bi)

e(IV)(ABSA)2(ads)(Bi) + 4e → Se(0)(ABSA)2(ads)(Bi)

.1.4. The effect of CETAB concentrationThe cationic surfactant (CETAB) can enhance the stripping sen-

itivity of the Se(IV)–ABSA complex at a bismuth film, due to theissociation of CETAB (a quaternary ammonium salt) into a simplenion and cation [30]. The positively charged cations transfer to the

Fig. 4. The effect of deposition potential and time on

Fig. 3. The effect of CETAB concentration on the stripping current of Se(IV)–ABSAcomplex.

cathode and are concentrated on the surface of electrode in the pro-cess of preconcentration. This preconcentration induces a changein structure to create an electric double-layer, which results in eas-ier preconcentration of the Se(IV)–ABSA complex on the surface ofthe electrode. Therefore, the stripping peak current is increased.

In the experiment, when the concentration of CETAB was var-ied from 0 to 2.19 × 10−5 mol/L with 10 �g/L Se(IV) (as shownin Fig. 3), the peak currents increased with the concentration ofCETAB in the low concentration and kept stable in the range from2.74 × 10−6 to 1.64 × 10−5 mol/L, then started to decline at concen-trations in excess of 1.64 × 10−5 mol/L. Therefore, the concentration1.37 × 10−5 mol/L was used in the experiment.

3.1.5. The effect of deposition potential and timeThe optimum electrochemical conditions for the preconcentra-

tion of Se(IV)–ABSA on the bismuth film electrode in 0.15 mol/L(pH2.9) HAc–NaAc + 1.0 × 10−4 mol/L ABSA + 1.37 × 10−5 mol/L CETABsupporting electrolyte were determined by varying the depositiontime and the potential of the bismuth film. The potential range of−0.3 to −0.8 V was chosen in the experiment with a deposition time

120 s. When the deposition potential shifts in the negative direc-tion, the peak current enhanced gradually in the range of −0.3 to−0.4 V. This peak reached the maximum at the potential of −0.4 V,then started to decrease (Fig. 4a); thus, the potential of −0.4 V wasused in the experiment.

the stripping current of Se(IV)–ABSA complex.

4720 Q. Zhang et al. / Electrochimica Acta 55 (2010) 4717–4721

Table 1Results for the determination of selenium in different samples and recovery rate.

Sample Certified value (�g/g) Found (�g/g)a RSD (%) Recovery

Found (�g/L) Addition (�g/L) Found (�g/L)a Recovery rate (%)

Multi-vitamin tablets 20.1 19.4 1.8 9.7 6 15.5 96.110 19.5 97.0

b 0.60 0.58 2.6 3.0 10 12.9 98.7

) (GB

scit1tta

3

po2e(clam

3

coadiE

F2

Human hair

a n = 3.b From China National Research Center for Certified Reference Materials (NRCCRM

The deposition time was then varied between 0 and 300 s. Fig. 4bhows the influence of deposition time (0–300 s) vs. stripping peakurrent of the Se(IV)–ABSA complex with stirring. The peak currentncreased with the deposition time from 0 to 120 s, indicating thathe amount of the deposited Se(IV)–ABSA complex maximizes at20 s when the concentration of selenium is 10 �g/L. However, inhe real sample analysis, due to the low concentration of the Se-ion,he favorable deposition time should be delayed; in our real samplenalysis, the deposition time was delayed to 300 s.

.2. Linear range, detection limit and reproducibility

Under the above experimental conditions, a sensitive strippingeak of the Se(IV)–ABSA complex was observed at the potentialf −0.76 V, for an deposition time of 300 s; the linear range of–30 �g/L was established (as shown in Fig. 5), the linear regressionquation was i = 0.1978c + 0.4476, where i and c are peak current�A) and Se(IV) concentration (�g/L), and the linear correlationoefficient (r) was 0.9985. The voltammograms obtained for theowest concentration with a deposition time of 300 s was 0.1 �g/L,nd the relative standard deviation from the eight parallel deter-ination of Se(IV) at 10 �g/L was 4.91%.

.3. The interference

When 8% error is allowed, many materials, i.e., a 1000-fold con-entration of Cl− and SO4

2−, 500-fold of K+, Mg2+ and Co2+, 300-fold

f Mn2+, 200-fold of Na+, 100-fold of Zn2+, 10-fold of Ni2+, Ca2+, Hg2+

nd Cr3+, 5-fold of Fe3+ and 2-fold of As3+, do not interfere in theetermination of 10 �g/L Se(IV). The 2-fold of Cu2+, Pb2+ and Cd2+

nterfere, but this interference can be eliminated by the addition ofDTA.

ig. 5. The differential pulse voltammograms of Se(IV): (a–j) 0, 2, 4, 6, 8, 10, 15, 20,5, 30 �g/L and the calibration graph.

[[[[

[

[[

[[

[

20 23.1 100.3

W07601).

3.4. Sample measurement

In our experiment, the developed method was applied to twokind of real samples: multi-vitamin tablets and human hair. Thesamples were digested according to the procedure described inthe experimental section, then the sample solutions were used forthe analysis of selenium. The results are listed in Table 1, and therecovery rate is from 96.1 to 100.3%.

4. Conclusion

In this work, the determination of selenium by adsorptivestripping voltammetry on a bismuth film electrode has beeninvestigated. The bismuth film electrode is characterized by highsensitivity, good reproducibility, a more easily renewed electrodesurface and is more “environmentally-friendly” than mercury. Thisstudy has revealed the possibility of considering the bismuth filmelectrode as a good alternative to the mercury-based electrodes.The proposed method can be applied to the determination of traceselenium in a real sample.

Acknowledgements

This work was supported by the Dean Foundation of ChineseAcademy of Inspection and Quarantine (2008JK013), the NationalNatural Science Foundation of China (20775088) and the Foun-dation of State Key Laboratory of Environmental Chemistry andEcotoxicology, Research Center for Eco-Environmental Sciences,Chinese Academy of Sciences (KF2008-06).

References

[1] M. Ochsenkuhn-Petropoulou, F. Tsopelas, Anal. Chim. Acta 467 (2002) 167.[2] M. Ashournia, A. Ali akbar, J. Hazard. Mater. 168 (2009) 542.[3] V. Negretti de Bratter, S. Recknagel, D. Gawlik, Fresen. J. Anal. Chem. 353 (1995)

137.[4] N. Maleki, A. Safavi, M.M. Doroodmand, Talanta 66 (2005) 858.[5] J. Stripeikis, M. Tudino, O. Troccoli, R. Wuilloud, R. Olsina, L. Martinez, Spec-

trochim. Acta Part B: Atom. Spectrosc. 56 (2001) 93.[6] M. Díaz-Somoano, M.A. López-Antón, M.R. Martínez-Tarazona, Fuel 83 (2004)

231.[7] B. Izgi, S. Gucer, R. Jacimovic, Food Chem. 99 (2006) 630.[8] B. Sun, M. Macka, P.R. Haddad, J. Chromatogr. A 1039 (2004) 201.[9] P.V. Dael, D. Barclay, K. Longet, S. Metairon, L.B. Fay, J. Chromatogr. B: Biomed.

Sci. Appl. 715 (1998) 341.10] A. Chatterjee, H. Tao, Y. Shibata, M. Morita, J. Chromatogr. A 997 (2003) 249.11] D.E. Mays, A. Hussam, Anal. Chim. Acta 646 (2009) 6.12] M. Pesavento, G. Alberti, R. Biesuz, Anal. Chim. Acta 631 (2009) 129.13] P. Rach, H. Seiler, Polarography and Voltammetry in Trace Analysis, Huthig,

Heidellberg, 1987.14] E.A. Hutton, B. Ogorevc, S.B. Hocevar, M.R. Smyth, Anal. Chim. Acta 557 (2006)

57.15] J. Wang, B. Tian, Anal. Chem. 65 (1993) 1529.

16] J. Wang, S.B. Hocevar, R.P. Deo, B. Ogorevc, Electrochem. Commun. 3 (2001)

352.17] M.A. Nolan, S.P. Kounaves, Anal. Chem. 71 (1999) 3567.18] E.A. Hutton, B. Ogorevc, S.B. Hocevar, F. Weldon, M.R. Smyth, J. Wang, Elec-

trochem. Commun. 3 (2001) 707.19] J. Wang, J. Lu, Electrochem. Commun. 2 (2000) 390.

mica A

[

[[[[

[

Q. Zhang et al. / Electrochi

20] E.A. Hutton, S.B. Hocevar, L. Mauko, B. Ogorevc, Anal. Chim. Acta 580 (2006)

244.

21] A. Królicka, A. Bobrowski, Electrochem. Commun. 6 (2004) 99.22] G. Kefala, A. Economou, A. Voulgaropoulos, Electroanalysis 18 (2006) 223.23] G. Jarzabek, Z. Kublik, Anal. Chim. Acta 143 (1982) 121.24] F.A. Bertolino, A.J. Torriero, E. Salinas, R. Olsina, L.D. Martinez, J. Raba, Anal.

Chim. Acta 572 (2006) 32.

[[[[[

cta 55 (2010) 4717–4721 4721

25] C.F. Pereira, F.B. Gonzaga, A.M. Guarita-Santos, J.R. Souza De, Talanta 69 (2006)

877.

26] M. Ashournia, A. Aliakbar, J. Hazard. Mater. 174 (2010) 788.27] M. Ashournia, A. Aliakbar, J. Hazard. Mater. 168 (2009) 542.28] F. Li, J. Chen, J. Nanjing Univ. Chem. Technol. 22 (2000) 52.29] C. Sun, J. Wang, X. Li, H. Yin, W. Jin, Chin. J. Anal. Chem. 19 (1991) 139.30] Q.Z. Si, J.M. Wu, J. Chen, J. Zhejiang Univ. (Sci. Ed.) 22 (1995) 282.